Question Regarding Armstrong-type Oscillators

[Fischer777]

So a little background: I am trying to construct a superhet radio with an oscillator producing a harmonic-rich output that can be electrically coupled to a mixer, the idea being one of the harmonics can heterodyne with the RF signal. After doing research regarding the construction of Armstrong-type tube oscillators I have found that the majority of designs (or at least those using a tuning capacitor where one side is earthed) follow one of the three topographies shown in the attached picture. I am wondering what the advantages/disadvantages to each design are. My source calls A the Meissner variant and states it's practical purposes are limited, B seems to be the most commonly used variant, and C appears to be used in the mixers of older superheterodyne radios (suggesting maybe it's more resistant to an electrically-coupled load?).

Thanks in advance.

 

[the_emi_guy]

Hey, interesting project!

Is this for AM band or FM band (or some other band)? Receiving or transmitting?

AM receivers generally use(d) pentagrid converters like the 12BE6 (in lieu of separate BFO and mixer). Is there a reason you want separate BFO?

Maybe there is a real old-timer here that has actual design experience with these... (I tinkered with them as a teenager). Look around for some old reference material like ARRL handbooks from the 40s and 50s, I bet you can purchase them on Amazon. I used to have a wonderful old RCA manual, but its long gone now.

Here is one that is available online:

http://www.ax84.com/static/rdh4/chapte24.pdf

 

[Fischer777]

It's for a broadcast band AM/FM receiver. I'm using a separate BFO as I've read that even at the lower frequency end things like space charge and capacitance between the various grids in pentagrid tubes can destabilize the frequency of the oscillator. I have a few books on tube oscillator construction, but none of them give me much information regarding Armstrong oscillator design. Thanks for the PDF, I'll be sure to check it out!

 

[skeptic2]

My first reaction is to favor B because the cathode is grounded. My experience with transistors, not tubes, indicates that impedance in the emitter circuit can cause instabilities which produce spurious radiation. But perhaps that is what you want.

 

[Fischer777]

Okay, thanks guys. I'll do more research regarding the advantages/disadvantages of an output taken from the plate and from the cathode. One more question though: is there a difference in stability or output when the resonant circuit is attached to the grid as appose to the output?

 

[skeptic2]

Probably not a significant difference. I would be concerned about a varying load impedance from the mixer or a signal feeding back from the mixer to the oscillator. Both are probably remote. However, if the tuned circuit is on the grid, it will probably be more sensitive to these effects as the grid circuit has a higher impedance and a lower signal voltage.

How are you planning to connect the local oscillator to the mixer? There are some circuits on this site especially towards the bottom.

http://www.rfcafe.com/references/qst/practical-design-of-mixer-converter-circuits-feb-1941-qst.htm

 

[Fischer777]

I was thinking about coupling the anode of the oscillator to the screen grid of the mixer; I've read that the oscillator would need to be strong to impress a great enough signal voltage across the screen grid to sufficiently vary the amplitude, but other than that it works well. I'm not sure if such a method of coupling would significantly isolate the oscillator from variations in impedance or from signal feedback so once I get the oscillator to work (which is why I asked about the different layouts of the Armstrong; I can't seem to find one that works) I'll experiment around different methods of coupling.

Thanks for the suggestions and the link!

 

[the_emi_guy]

I'm guessing that 108MHz with tubes may be a challenge. It's obviously possible, they were for sale.

Years ago I attempted something like this and had a heck of a time with the frequency shifting significantly as I moved my hand away from variable cap tuning knob, and I was only attempting 10MHz. Then again I didn't really know what I was doing at the time.

 

[jim hardy]

That Tube Manual link posted by EMIguy is great info, it's where I would start.
Plenty of them on Ebay.
http://www.ebay.com/itm/rca-receiving-tube-manual-/321172147772?pt=LH_DefaultDomain_0&hash=item4ac75a0e3c

Look in thrift stores and yard sales for an old AM/FM tube radio to study or restore. A non-working one will be less expensive.
They are easy to work on - you can see all the parts, and there's plenty of antique radio sites .
http://antiqueradio.org/radweb.htm
http://www.justradios.com/links.html
Or just find some model numbers and look for service manuals. Sams Photofacts have great schematics with expected voltage readings.

FM tube radios were harder to work on because of the higher frequency. First time through I'd make two different tuners.


You are aware that stuff is now all synthesized in IC's ?

http://www.ti.com/solution/digital_radio

 

[Fischer777]

I'm guessing that 108MHz with tubes may be a challenge. It's obviously possible, they were for sale.

It's a 6KE8 tube, which is designed for use as a television oscillator/mixer. Also, because I'm using the third-order harmonics from the oscillator, I won't need to design one that has to work at such high frequencies. The problem isn't so much I've got the wrong frequency as it is the oscillator just doesn't work. I suspect it may be because the tube itself is quite far removed from where the LC circuit is, so I'm going to try using coaxial cable for the connections to see if that clears anything up.

That Tube Manual link posted by EMIguy is great info, it's where I would start.
Plenty of them on Ebay.

I've got a book on the construction of oscillators (covering both tube and transistor construction methods) and another on the construction of FM transmitters/receivers (which focuses entirely on tubes). I've also been reading the links by EMIguy and Skeptic2.

Look in thrift stores and yard sales for an old AM/FM tube radio to study or restore. A non-working one will be less expensive.

I've repaired 3 old tube radios (one of which was an AM/FM) and a tube television, which is one of the reasons I wanted to build my own; it seemed to be the next step.

FM tube radios were harder to work on because of the higher frequency. First time through I'd make two different tuners.

The intermediate frequency is 20 MHz, so the oscillator needs to generate a frequency from 67-88 MHz, and because it's using third-order harmonics from (what I hope is) a class-c oscillator, the fundamental frequency will at it's highest be 29.3 MHz. This hope is that this arrangement will make construction of the oscillator/mixer go more smoothly.

You are aware that stuff is now all synthesized in IC's ?

Yes, but analog stations are more numerous than digital stations where I am (at least for now).

 

[jim hardy]

I've repaired 3 old tube radios (one of which was an AM/FM) and a tube television, which is one of the reasons I wanted to build my own; it seemed to be the next step.

I applaud your enthusiasm and curiosity - Have fun !

My interest never went much beyond fixing them. And some vacuum tube audio - I still have a push-pull 6V6 Ampex outfit... a 1930's Zenith wood table radio(Loctal tubes) and a 1940-ish Airways. I let two old Armstrong consoles get away , they were in his old FM band but the cabinets were nice. Ahh almost forgot one ancient battery set with #30 triodes...


Best wishes !

 

[Baluncore]

I would recommend that you get a copy of the 4th edition, “Radiotron Designer's Handbook”.
Google to find a PDF format.
It was up-to-date in 1952, just before the advent of the transistor. Over 1400 pages of vacuum tube practice in audio and RF.

 

[Fischer777]

Thanks both of you guys! It seems like between all the links and PDFs I've got a lot of reading to do. I look forward to seeing if I can get this to work.